In the seventh year of the French Republic (1799 to the rest of the world), some peasants of Tarn and Aveyron, in southern France, encountered a naked boy scavenging alone in their fields and forests. He did not speak, and seemed not to understand any French. At first he ran away from other humans. More than once he was captured and brought to town; each time, he escaped. Later, the boy became familiar to the mountain farmers. He would appear in their houses during the day to be fed, and then disappear again every night. Some claimed he moved unusually fast, on four limbs. Others claimed he rejected meat, and inferred from this that human beings are not naturally carnivorous. One night in 1800, while he was taking shelter from a storm, the boy was captured for good. His family and past were unknown and became the topics of intense speculation. Had he been abandoned at birth? Had he intentionally escaped from brutal parents? Because he did not understand language, he was initially—but inaccurately—assumed to be deaf. Eventually he was transferred to the care of Abbé Sicard, the head of the Institute for Deaf Mutes in Paris, and to the protection and investigation of the Society of Observers of Man.

In Paris, the “wild boy,” now named Victor, was initially an object of immense curiosity, but the public quickly lost interest. The first team of philosopher-observers from the Society despaired of any progress (concluding that there was “the greatest degree of probability” that the boy had been born either an idiot or insane) and gave him up. Then a new teacher emerged: the physician Jean-Marc-Gaspard Itard took over Victor’s care and worked with him daily for over two years. Using a combination of food rewards and physical punishments, Itard forced Victor through set after set of newly devised linguistic exercises. Eventually, Victor did learn some basic signs, but, critically, he never learned to speak. Itard gave up in 1806. From then until his death in 1828, Victor lived in anonymity with a guardian, Mme. Guerin. Itard, on the other hand, remained prominent throughout his lifetime and was later remembered as a pioneering scientist, psychotherapist, and teacher of disabled children.

Scientists’ thoughts returned to Victor and Itard in 1970, when a radically isolated girl was discovered in Los Angeles. “Genie” was 13 at the time but unable to walk or talk. She had, it seemed, spent most of her life in a room of her parents’ house. Initially scientists saw her as an extraordinary opportunity to study and teach a “wild child,” and she was taken to live in the home of one of the psychologists. She made very little progress, though, learning just a handful of words over the next four years. In 1975 the federal grant that funded her care was not renewed. For the next few years, writes the historian Adriana S. Benzaquén, “Genie lived in a succession of foster homes; she was mistreated and physically abused again; she lost the few skills she had learned . . . and she stopped speaking altogether.”

The life stories of Genie and Victor fit a pattern established over centuries of scientific and philosophical “encounters with wild children”—the title of Benzaquén’s new book. Benzaquén illustrates and seeks to make sense of this pattern: the extreme high hopes, proportionate disillusionments and dubious moral choices that have cycled through societies’ responses to children deemed “wild.” In other words, she tells the tales of our encounters and only secondarily the tales of the children encountered. Ultimately, her aim is to expose the intellectually and ethically suspect decisions made by those involved in the construction of the tales. Although Benzaquén sometimes lapses into dense academic jargon, especially in the first two chapters, her book is a compelling read. We are both caught up in the fascination of the stories and forced to confront this fascination, to regard with suspicion “the search for the truth about wild children (and the truth in wild children)” that continues to this day.

In every generation, the idea of a child growing up in isolation from society provokes deep and persistent questions about what it means to be human. Which parts of ourselves are determined by biology and which by culture? To what extent is language innate? Can moral instincts develop without instruction? Is even walking on two feet dependent on cultural transmission? Some philosophers have argued that society contaminates human beings, others that it ennobles us. For both sides, the way to resolve these questions, to “definitively reveal ourselves to ourselves,” has seemed in equal measures tantalizing and taboo: the forbidden experiment, Benzaquén calls it.

“A prince could do a beautiful experiment,” wrote Montesquieu. “Raise three or four children like animals, with goats or with deaf-mute nurses. They would make a language for themselves. Examine this language. See nature in itself, and freed from the prejudices of education; learn from them, after they are instructed, what they had thought; exercise their mind by giving them all the things necessary to invent; finally, write the history of the experiment.” Centuries later, the secret appeal of such an experiment—if slightly updated—is unabated. Wild children intrigue and enthrall because they seem to offer a morally permissible version of the forbidden experiment, one whose initial conditions are created not by cruel scientists but by cruel parents or cruel accident. Historically, though, this natural forbidden experiment has invariably failed to deliver. The scientists, philosophers, and pedagogues involved have left records of disappointment. The children themselves have died young, sunk into anonymity, or been abandoned to further neglect and abuse. The grand questions about human nature remain unanswered.

Three patterns of failure recur. In the first, the wild child is never sufficiently rehabilitated to serve as a witness, perhaps because the consequences of linguistic, emotional, and social deprivation are too devastating. Instruction fails, so the observers can never “learn from them, after they are instructed, what they had thought.”

In the second pattern, rehabilitation works all too well. Instruction destroys the unique wildness of the child. The former wild child can talk about his or her life experiences but has become a suspect witness, just as contaminated by society as the rest of us. (Some scientists anticipated this quandary, prompting sentiments that Benzaquén finds unsavory. She quotes Harlan Lane, for example, contemplating the 20th-century discovery of John of Burundi, thought to have been raised by monkeys in the Ugandan jungle: “All this teaching the boy is well and good, but it is obliterating the traces of life in the wild and is destroying his value as a scientific discovery.”)

Whether instruction succeeds or fails, the true wildness or isolation of the child inevitably comes into doubt, amounting to a third kind of failure. In general, almost nothing is known about a putative wild child’s life either before or during the period of isolation. Either there are no witnesses to the child’s life pre-capture or the few existing witnesses contradict themselves and are in any case not disinterested. So far, the claim that any specific child has survived for more than a few weeks away from human society has never been proved. As a result, the consequences of isolation per se are almost impossible to determine or defend. If a child is responsive to instruction, skeptics charge that the child was never truly isolated. If the child cannot be instructed, they (or the disappointed scientists themselves) conclude that the child is a “congenital idiot,” that incurable language delay or emotional trauma were inevitable in this child from birth, and not the consequence of isolation.

In spite of this record of failure, each successive generation has faced its own encounters with wild children with renewed high hopes. Why? Benzaquén’s answer is the one really disappointing part of her book. Largely, she assigns the blame to the blind and hubristic ambition of scientists seeking personal fame. About Genie, Benzaquén writes: “For people in general, she was an object of pity; for scientists, she was an object of knowledge…What professional and personal rewards would Genie not have in store for whoever was there, ready to grab them?” About the scientists who set out to study John of Burundi: “Their words and actions betrayed the over-confidence of the Western scientific researcher (and the white American male) storming into the unsuspecting Third World.”

Scientists do have personal ambitions, it’s true, and like most human beings, scientists can be racist and hypocritical and can make bad moral choices in complex situations. Benzaquén may be right that in the treatment of wild children bad moral choices have been all too common, and poor scientific judgment has certainly been rife. (Alarmingly, as Benzaquén’s book was going to press, in March 2006, the BBC announced the discovery of five siblings in rural Turkey who walk quadrupedally, claiming that this family never “made the leap” to a bipedal gait and serves as a “living example of how our ancestors walked.” The scientists in this case simply ignored the fact that the five siblings are the first and only generation of their family who walk quadrupedally. The first group of Turkish scientists even announced that the family represented an evolutionary “lost link,” exhibiting only primitive language—until it was subsequently revealed that the family spoke Kurdish.)

And yet, simply vilifying the scientists is too easy. There are forces more interesting than personal ambition at work in these successive failures. The progress of science in the last three centuries has been so remarkable partly because scientists are trained to regard the failures of the preceding generations as non-definitive, as marking a space for improvement and innovation. As scientific tools and techniques improve and bodies of knowledge expand, we see what was previously invisible. By far the most common kind of failure in the history of science has been this temporary kind, the kind that can be overcome by the next generation.

Because of this generally successful tradition, each failure to learn from a wild child in the past may be just a technical failure. Each new generation rests its hopes on “modern” methods for teaching language, or for measuring cognition in the absence of language, or for assessing or improving emotional functions. As Harlan Lane wrote, comparing the scientists who would study John of Burundi in the late 20th century to Itard, Victor of Aveyron’s teacher in the early 18th: “How much more could we discover about what it means to grow up in society from this terrible experiment of nature, which chance had designed and which science could exploit? And how much more could we contribute to the education of handicapped children everywhere by undertaking the training of this latest, and perhaps last, wild child, raised in the forests utterly cut off from society?” These continuously renewed hopes spring most fundamentally not from the weaknesses of individual scientists but from one of the greatest strengths of science as a whole.

But here’s the catch: the forbidden experiment may belong to a smaller group of experimental problems that persistently seem meaningful but are not. Intuitively, we expect that while human nature interacts with human society in a typical child’s development, the natural and the social are in principle independent and distinguishable. If this intuition is wrong, the forbidden experiment is incoherent. In fact, the social and the natural may be irretrievably entangled in development. In part this is because a social environment that includes other human beings is inevitably more natural for a human infant than any wholly artificial environment that could be constructed to replace it. Even the unfolding of innately determined human traits relies on a social environment. For example, virtually every human infant is exposed to a language and learns it; an infant who was never exposed to any language could not possibly speak one. Yet it is the children who do learn a language—through social interactions—who illustrate the natural human capacity.

If the forbidden experiment is indefensible not just because it is immoral but because it is incoherent, Benzaquén has—despite her sometimes dismissive attitude—done modern scientists an important service. Her book teaches us about failures in our history to which we must pay more attention than usual because these failures cannot simply be overcome.

In her final paragraphs, Benzaquén extends her moral condemnation from the specific scientists who have figured in the lives of wild children to everyone who studies, teaches, or theorizes about children. All adults who care for children face a key moral challenge, she says: to “reconcile the conflicting demands, on the one hand, to approach the child as another subject whose integrity, separateness, and freedom ought to be maintained, and on the other, to care for the child, intervene, interfere, educate, mould, change.” Benzaquén’s charge is that by making children the object of study, “experts” on childhood actually oppress children and undermine their agency by “turn[ing] a moral question into a scientific one.”

I am moved by the moral challenge that Benzaquén describes here but strongly disagree with her conclusions about the sciences of childhood. In fact, even the stories in her own book contradict her pessimistic assessment. Over the past three centuries, the language impairments of wild children have often been contrasted with those of deaf or developmentally disabled children. As a result, contemporary expectations for the lives of deaf and disabled children make regular appearances in this history. In 1801, when Victor of Aveyron was first brought to Paris, for example, the common wisdom was that deaf children were incapable of thought. Victor was initially under the care of Abbé Sicard, whose new school for the deaf was considered a revolutionary experiment. The successes of his students brought Sicard immense fame and helped win recognition of the now unremarkable fact that deaf children are fully human.

The trajectory of the developmental sciences more generally has proceeded in the same direction, continuously increasing our appreciation of young children as worthy of interest and respect and as conceptual thinkers in their own right. Itard struggled and finally failed “to lead [Victor] to the use of speech by means of imitation and ‘the urgent law of need.’” What we now know is that learning the meaning of words is one of children’s most striking accomplishments. By the time an English speaker is 17 years old, she knows, on average, about 60,000 words, more than ten for every day of her life.

Even more interesting, though, is how children are learning all these words. The simple way to learn the meaning of words, one might imagine, is to hear an unfamiliar sequence of sounds (“whisk”) and associate that word with a concurrently visible object. In fact, children’s word learning is much more sophisticated. In one experiment, scientists give an 18-month-old child an interesting novel object (e.g. an eggcup). Once the child is absorbed with the object in her hand, the experimenter looks into a box and says “Oh, a whisk!” If children just associated novel sounds and novel objects, then the child should learn (erroneously) that the object in her own hand is a whisk. But that’s not what children do. Instead, they look up at the scientist, follow her gaze into the box, and conclude that a “whisk” is whatever is in the box. Asked later to “find me a whisk,” these children pick the whisk, not the eggcup. In short, what we have learned from studies of very young children is that they are already making rich and sophisticated inferences about the object to which the adult intends this new word to refer.

In all, the tales of wild children are striking and instructive but atypical. The history of the developmental sciences is not merely a history of failures. Through experiments that are not forbidden, we do, slowly, reveal ourselves to ourselves. Learning from and about childhood can be both a scientific endeavor and a moral one.